444 T. S. Hall — Graplolite Rocks, Victoria, Australia. 



persisted, while the commonest associate is Didymograptus caducens, 

 which is represented by the large variety characteristic of the higher 

 beds of the Castlemaiue Series. In Canada the only species given by 

 Ami^ as associated with it are Rouvilligraptus Bichardnoni and Gonio- 

 graptiis Thnreaid. The former species has not been recognized in 

 Australia, while the latter is confined to a lower horizon, not passing 

 up into the Castlemaine Series. In the Skiddaw Slates it would 

 also appear that the position of Loganograptus is lower in the series, 

 for Miss G. L. Elles,*in her table of suggested phylogeny, puts down 

 Diclwgraptus oclohrachiatns and Tetragrnptns qundribrachintus as de- 

 rived from Loganograptus Logani. Genealogical trees are notoriously 

 difficult to draw up in biology where there is great danger of 

 confusing the different collateral branches. The question suggests 

 itself as to what is the real order of the appearance of these forms. 

 Migration, one would think, must produce apparent inversions in 

 widely separated regions, and it will be evident that several such 

 inversions actually do occur in the case of our graptolites, or, in 

 other words, that we find in Victoria forms associated which are not 

 found together in the Northern Hemisphere, and vice versa. That 

 the general phylogeny of the graptolites may be traced out on the 

 lines indicated by Nicholson and Marr,^ and so ably elaborated 

 by Miss Elles, there can be but little question, but the task of 

 elaborating the details is complicated when we take into con- 

 sideration the possibility that the succession of life in any particular 

 set of beds may not be the true one. The general succession of the 

 graptolites in Australia and the Northern Hemisphere is the same, 

 but there are undoubtedly differences in details. 



The Castlemaine Series has been best studied in the district 

 whence its name is taken, and it there includes, as far as I have seen, 

 all the beds above the Bendigo Series, which is but slightly exposed 

 on the main anticline which passes through Chewton. 



The group includes rocks occurring at Castlemaine, a large area in 

 the Bacchus Marsh and Coimaidai Districts,* Gisborne, and perhaps 

 a considerable part of the intervening country, as well as an area in 

 the Whipstick, north of Bendigo. 



The fauna includes Didymograptus bifidus, J. Hall ; D. Murchisoni, 

 Beck ; D. cf. decens, Tqt. ; D. extensus, J. Hall ; D. caduceus, Salter ; 

 Tetragraptiis serra, J. Hall ; T. qiiadribrachiatus, J. Hall ; T. pro- 

 oectus, n.sp. (v. infra); Dichograptus octobrachiatus, J. Hall; D. 

 octonarius, J. Hall ; Glonograptus, sp. ; Phyllograptus typus, J. Hall ; 

 P. angustifolius, J. 'HslU; Loganograptus Logani, J. Hall; Gonio- 

 graptus macer, n.sp. (v. infra) ; Dendrograptvs, sp. ; Diplograptus, 

 spp. ; Climacograptus, sp. ; and perhaps Trigonograptus, as well as 

 many forms as yet unidentified. 



4. Darriwill Series. — The only locality at which these rocks are 

 known is in the parish of Darriwill, where graptolites were found by 



1 Geol. and Nat. Hist. Surv. Canada, new ser., Ann. Kep., vol. iii, pt. 2, K, p. 116. 



2 Q.J.G.S., vol. liv (1898), p. 537. 



3 Geol. Mag., 1895, pp. 529-o39. 



* Proc. Eoy. Soc. Victoria, new ser., vol. x, 1897 (1898), pp. 202-3. 



